324

OLD ROADS IN CENTRAL HANTSBY C. F. C. HAWKES.

Although many of the roads to be found in this countryare of very great age, yet none of them are older than the earliestsettlements. Villages and camps were not established, along theline of already existing roads, but roads grew up as links betweensettlements. Before the Saxon invasion all the villages wereoh the uplands, and hence the roads followed the lines of theridges from one to another. This was the best way for severalreasons:—It was the driest, and generally avoided streams, andit lay over open ground; moreover, the safest route for travel

. is always that with the widest view. But the principal reasonwas that the uplands were the centres of population.

The oldest prehistoric roads are the long trade routes stretch-ing right across the country, which form the skeleton of the wholesystem: they keep, wherever possible, to the high ridge topsor just on the south brow, and where a ridge has to be crossedit is usually at a saddle with an easy gradient: for the crossingof a river valley, they generally descend by some spur with a gentle slope, and always choose a place as high up the river aspossible, where the difficult wet valley is narrow. But laterthere grew up parallel roads beside them, which kept to thelower slopes of the ridges, and were used in dry summer weather,when travellers were not forced to get right to the crest of a ridge to find a passable road. These summerways can be followedalongside of most of the big ridgeways: and it is noticeable thatwhen a river is to be forded they usually come in and join theparent ridgeway for the crossing, leaving it when it climbs thehill on the other side. The characteristic of the ridgeway isthat it seldom leaves the highest ground.- The summerway isgenerally to be found somewhere near, but not at the bottomof the least indented slope of the ridge. Where both slopes areequally straight (like those of the Hog's Back), it will choose thedrier southern side.

The minor ways of the pre-Saxon system are now few, andmost of them are suspect of being really Saxon, or later still;for at the Saxon settlement an entirely new system of valley-habitation was introduced, with new roads and fields which dis-placed the old. The Saxon ways grew up in the valleys to connectthe Saxon villages, and are quite distinct from the British uplandroads. Occasionally, however, Saxon villages are found estab-lished along a summerway, which may have already beenin existence.

In Central Hants the main ridgeways run roughly east andwest, with connecting links between them, and branches southto the coast.

(i.) The Harroway was the most important of the Hamp-shire ridgeways, and is a part of the great road which stretched

325

right along the backbone of Southern England from Cornwall•to Kent. From Salisbury Plain it came along the high groundnorth of the Test valley, through Weyhill, and, skirting Hurst-bourne Park, it passed just north of Whitchurch, parallel tothe L.S.W.R. At Oakley, on the low saddle between the Testand Lpddon basins, it joins a road coming south-east from theBurghclere downs, and must originally have gone on towardsthe tumuli north of Kempshott House. Beyond the Winchester-Basingstoke road a lane marks its route through Farleigh Wallopand Ellisfield: further, it must have passed through Nash'sGreen and Weston Common to the Golden Pot Inn, on theBasingstoke-Alton road, and then it climbed the ridge and wentpast Sutton Common, and the Penley Copse and Barley Poundearthworks, into the Wey valley at Farnham.

(ii.) But the summerway, which leaves it at Oakley, andpasses over Battle Down, and between Basingstoke and Cliddes-den, along the edge of Hackwood Park, through Five LanesEnd and Long Sutton, to join it again near Well, seems graduallyto have absorbed traffic till it superseded its parent, which isnow consequently' in a fragmentary condition. We shall meetother instances of this.

(iii.) First branch of the Harroway. This is a link between- the Harroway and the South Hants ridgeway, and, leaving the

former not far west of Cholderton, it passes the Quarley Hilland Danebury Camps. It crosses the Test at Longstock, and

' can be traced as a lane up to Upper Sandy Down Farm, nearStockbridge. Its course is uncertain until -it re-appears south-west of Woolbury Ring as a deep hollow way, which continuesjust across the Winchester-Stockbridge road. After this thereare no traces : presumably it went on to join road (iv), somewhereeast of Upper Somborne.

(iv.) The second branch of the Harroway leaves it nearFinkley, and goes west and south. After crossing the Test nearFullerton Junction, it is traceable as a wide lane past the " WhiteH a r t " and over Windmill Hill. South of the Winchester-Stockbridge road it is narrower" and deeper, and just north ofthe Somborne-Crawley road hollow tracks are visible. Furthersouth the lane becomes a very wide, green road, and turns east;as a lane, it crosses the ridge between the Test and Itchen vaileysnear its lowest point at Ham Green, where it has some hollowtracks, and can be followed as far as Weeke. It must have joinedthe South Hants ridgeway in Winchester.

(v.) The Lunway has been thought not to be a true" ridge-way, but it has all the characteristics of one, and I have foundno reason for doubting its authenticity. West of the Test I havenot explored it, but I imagine it to have come from the OldSarum region by way of Lobscombe Corner, and to have crossedthe Test at or near Stockbridge. At Woolbury Ring, which hasbeen wrongly thought to be its western terminus, it is not visible,

326

but it probably followed the north edge of the wood by Phillips'Heath Farm, and thence eastward it is visible as a lane alongthe ridge that bounds the Micheldever valley on the south.There is a slight gap. by the Winchester-Mildenhall Roman road,but in Isaac Taylor's County Map of 1755 a continuous courseis shown. Further on it is distinct past Worthy Down Race-course and the British village thereon, and, continuing just southof Waller's Ash Tunnel, it reaches the Winchester-Basingstokeroad at Lunways Inn, though not visible for the last fewhundred yards. Here it is joined by a Saxon drove road fromAndover to Alresford, which pursues a parallel course on itsnorth side over Worthy Down, after coming south throughBarton Stacey from Andover. As the real Lunway is to thesouth of the ridge-crest here, and the drove practically on it,some confusion has arisen, but, as there is no point of contactat the west (or Crawley Down) end of the parallels, it is clearthat they are quite distinct, the southern one being the realridgeway, the northern one Saxon. On the east of LunwaysInn they are united for some way, and then,, at the south-eastcorner of Micheldever Wood, there is a fork, one road goingnorth-east to the Candover, at Totford, the other south-east toAlresford, with a branch to Itctien Stoke. Whether the southerlyroad was a minor ridgeway before being used as a Saxon drove,I am not sure, but I suspect that the truth is as follows:—Inlate prehistoric or Romano-British times, a branch grew up fromthe main Lunway to join the road later known as the Pilgrims'Way at Itchen Stoke> and afterwards the drove took the samecourse as far as the tumuli on Itchen Stoke Down, and thenturned off to Alresford. This explains all three branches. I donot think the Lunway was connected with Oliver's Battery onAbbotstone Down on this side.

To return to the main line of the road, after crossing theCandover at Totford, a parish boundary alone marks its linefor a mile, and then a green lane takes it up on Bugmore Hill,whence a- short branch leads south to Oliver's Battery. Hollowtracks are to^be found in Godsfield Copse, and the way is con-tinued to Wield by a path along the south of Wield Wood andround Barton Copse. I t is dislocated by the villages of Wieldand Lower Wield, but there is a footpath between them, whichmay be a relic of it, and, further, a path and a lane mark it mostof the way along the ridge to Burkham House. I have notinvestigated it further, but the nature of the ground seems tonecessitate a winding course, and. it probably turned to the rightover Herriard Common and joined the Harroway near the Alton-Basingstoke railway-line, perhaps going beyond it over Prior'sHill to join its summerway at Five Lanes End.

(vi.) The Lunway does not throw off a summerway tillTotford is reached, but eastward of that it appears as a lane,and strikes north-east along the side of the ridge over Preston

327

Down ; it is then met by the modern road from Preston Candoverto Alton, and must have gone on to join the main Lunway againat Bagmore. It will be noticed that the ground covered bythe western portion of this summerway is not a very even slope,but it could not get anything better except by taking a veryzigzag course.

(vii.) At the corner of Barton Copse,, between Godsfieldand Wield, a branch ridgeway leaves the Lunway on the south—it is visible as a lane going south from the Wield-Medstead roadinto the woods. There it is lost for a little, but the modernroad from Heath Green to Hattingley and the direct path thenceto Medstead show its course: further east, I take it to haveturned south and followed the edge of Chawton Park Wood,and on across the railway to join another ridgeway, road (X),somewhere near Four Marks.

(viii.) The South Hants ridgeway is part of the big traderoute from South Wilts and Dorset, across Hants, and right alongthe South Downs. I t is described by Dr. Grundy in the "Archae-ological Journal" for March-December, 1918, but there areseveral points in which this account differs from his". Theridgeway runs eastward along the crest of Dean Hill into Hamp-shire, and crosses the Test at Kimbridge Mill, near Mottisfont.If probably followed the present road up to Michelmersh pastthe Bear and Ragged Staff, but its course through the villagecannot now be re-discovered.. Further on it is traceable as a lane running eastward on the south of Stubbs' Copse, and IsaacTaylor's map marks it running on along the ridge into ParnholtWood, where there are well-marked hollow tracks. Beyond,over Farley Mount Down, it is a wide lane, which joins the OldSarum-Winchester Roman road in West Woods. Originally theridgeway must have gone along the crest of Pitt Down, and thenthrough Crab Wood and the south part of the Royal WinchesterGolf Links into Winchester, but when all the traffic was absorbedby the Roman metalled road, that part of it fell into disuse, andit cannot now be traced.

Beyond Winchester it must have followed the same courseas the Roman and modern roads past St. Catherine's Hill to thecorner of Deacon Hill. This ascent is covered with hollow tracks,but all those now visible on the south of the modern road areSaxon and mediaeval; however, there are some on the northside- which are the remains of the old ridgeway. One set ofhollow tracks kept off the hill along the edge of the modernSewage Farm, and only scaled it just west of Chilcombe RifleRange, where they climb diagonally up the steep slope.

At the corner of Deacon Hill the Roman and modern roadsturn south, and the ridgeway is now lost, but the 6in. O.S. mapof 1897 shows a footpath continuing east to join " Steeple-chase ". lane, which runs north-east, south of the crest of the hill.This was the course of the ridgeway, and it is continued east

D

328

by hollow trails over Telegraph Hill up to the summit of Cheese-ioot Head. The north slope of Telegraph Hill is covered withdeep hollow tracks: those along the modern Winchester-Peters-field road must be comparatively recent, but the others, furtherwest, which disappear by Chilcombe Lime Works, are the remainsof a shorter, steeper, cut up from the river crossing at Winchesterto Telegraph Hill, avoiding the detour that the main ridgeway,makes to secure an easier gradient. There is a footpath runningW.N.W. from Chilcombe, that joins the road into Winchesterfrom Petersfield on the edge of the town, and in this we may see a relic of this branch. The steep and deep lane up from Chilcombeto the ridge-top behind the butts of the Rifle Range is certainlySaxon, and has nothing to do with any part of the ridgeway.There seems to be no evidence for Dr. Grundy's course roundto the north over Magdalen Hill. The route is steep, and nearthe New Inn is cut by a steep gully; moreover, there are nodefinite traces. At the top of Cheesefoot Head the hollow tracksare well marked in the first copse on the north of the presentroad, though the ground here is confused by the mediaeval London-Southampton highway1 (the modern Fawley Lane), which crossedthe hill from north-east to south-west.

The ridgeway's course on the eastern slope of CheesefootHead is followed by the modern road, but after the 500-footcontour is passed, the latter probably bore off to the north ; ina field about 50 yards south of it is a fold which may, perhaps,be a relic of old hollow tracks. Though cultivation has nowdestroyed it, it is marked by Isaac Taylor going south-east toLane End Down, where it is once more to be traced, broadeningout into well-marked hollow tracks which descend the hill pastsome.tumuli, and gradually contract towards the bottom.

Up the hill, from Lane End to Millbarrow Down, the ridge-way must coincide with the modern Winchester-Warnford road ; in several places it is quite deeply cut, and for-most of the wayit has a series of narrow fields and copses of uniform width alongits south side, which seems to show it was formerly much wider.On Millbarrow Down faint trails pass on both sides of the tumulus,to re-unite beyond. They were visible a little further on at thenorth edge of the road three years ago, but are now practicallyinvisible; then they disappear, and the line is taken up bythe lane leading from Lomer to Beacon Hill. What the coursemust have been in the short interval is shown by a parish boundary.

There is some doubt about the crossing of the Meon Valley.Dr, Grundy takes the road east and north from Lomer, pastthe tumulus on the north of Wheely Down, and then downa straight footpath into Warnford. But of the first portion ofthis course there are no traces, though there is no doubt thatthe wide lane leading up the smooth ridge from Warnford south-

1 See John Ogilby's Survey, 1682, and the map in Camden's " Britannia'," ed.Gibson, 1692.

329

eastwards on to Teglease Down represents part of the ridgeway.A second crossing of the valley is marked by the very deep hollowtracks down Beacon Hill, which, becoming shallower, curl southalong the 400-foot contour: their line is taken up, after oneintervening, field, by a lane leading to a ford not far north ofExton. A path (followed by a parish boundary) leads up fromthis to the Camp on Old Winchester Hill, through which it passesto join the way up from Warnford on Teglease Down. A middleroute from Beacon Hill, along Beacon Hill Lane to Shalford,and so up to Old Winchester Hill by Garden Hill Farm, is illattested, but, undoubtedly, both the first two were used. Dr.Grundy thinks the northern one is the original, although at thepresent day the other is more continuous. One might haveexpected a route from Beacon Hill to Warnford, but there seemsto be no trace of any. East of this the ridgeway is clearly trace-able as a wide, green road along the crest of Chidden Downand over Hyden and Tegdown Hills to Butser Hill, and thenit leaves the county along the -top of the South Downs, thoughI failed to trace it satisfactorily on Wor Down.

Unless the short cut from Winchester to Telegraph Hillbe so styled, this ridgeway has no summerway till Sussex isentered, but there it can be followed along the north edge of theSouth Downs from one end almost to the other. [See Dr. Grundyad loc]

(ix.) From Cheesefoot Head a branch of the South Hantsridgeway led south over Longwood Warren ; but the presenttrack heads off due south towards Old Down Farm, while theridgeway must have veered round to the south-west with theridge, and gone towards the Itchen by Hill Farm and the southside of Twyfofd Waterworks: of this nothing now remains.

(x.) 'On the summit of Lane End Down the South Hantsridgeway. has branches leading off at right angles on both sides.That on the south must have run down to the low ground some-where near Maxwell, but its course is only marked by a modernroad and a lane with a parish boundary almost as far as LongwoodFarm; beyond this it has disappeared. That on the north ismore interesting: its course north-east from Lane End Downis marked by a straight piece of the Bramdean-Owslebury road.When this turns off to the east, its line is continued by a lane,a hollow way along the edge of a field, and then again a lane.

, Near Hockley House the lane turns north, and the course of theridgeway is, I think, shown by the footpath leading north-eastacross the fields: when this becomes a lane, on approachingCheriton, there are steep banks on the left which seem to berelics of hollow ways. The road here is marked by Isaac Taylor.It is possible, however, that the footpath further south, byWestfield Farm, marks the true line. Anyhow, the ridgewaycrossed the Itchen at Cheriton, and its further course is shownby the path through the timber-yard and up into the fields beyond.

330

Thence a long trail, followed by a parish boundary, leads up theeven slope: it is deeply cut by three mediaeval tracks leadingsouth from Alresford, and becomes a wide, green road along theedge of Cheriton Wood. At the'top of the hill there is a slight gap,bridged only by the parish boundary, and then the ridgewaygoes as a big lane past Old Park Wood and the north edge ofBramdean Common.

I do hot think the lane on the north of Cheriton Woodhas anything to do with the real ridgeway. It ends abruptlyjust north of Cheriton battlefield, and the climb thither fromthe Itchen is much steeper and less likely than the route givenabove.

East of Bramdean Common all traces of this ridgewaydisappear for a good many miles; there are several old laneswhich descend the hill towards Ropley, but it must have goneon' along the crest. I suspect the lane past Merryfield Farmmarks it course, but it certainly did not turn off into the East-Tisted valley with the present lane (and parish boundary), though,as Dr. Grundy shows, this is an old one. The ridge turns northsward near Plairidell Farm, and the ridgeway must have continuedup to the top of the watershed near Four Marks, where it wouldbe joined by road (vii), but its precise route cannot now be found.Descending the further slope, it was probably coincident withthe modern Winchester-Alton road for a mile or so, and thenpassed across the railway to join the road which forms the northside of Alton But ts : but it must have borne off north fromthis by the Cripples' Home, and we next find a sure trace of itin the road running north from Anstey Manor. Marked by a modern lane, the ridgeway followed the comb of the ridge asit swings eastward, and joined the Harroway on the hill aboveBrockham Hill Farm.1

(xi.) As will be seen from the map, the object of this ridge-way was to connect the big port at Bitterne and its immediatehinterland with the Harroway and the eastern outlet affordedby the North Downs. The way is rather awkward, but musthave been important, and, to my mind, it is almost certain thatthe " Old Road " of Mr. Belloc first arose as its summerway,and later absorbed all its traffic. Certainly in Romano-Britishtimes, and perhaps before, this led to the disuse of the former,which is consequently fragmentary and difficult 'to trace to-day.The summerway is certainly rather a special case, as the natureof the ground keeps it further away than is usual from its parentridgeway, and it crosses from one side of the latter to the otheron the Medstead ridge, but I think that this view of the originof the " Old Road " is far better than to imagine that it grewup of itself without a previous ridgeway, or that it is merely

1 The " Maulth Way," mentioned in Dr. Williams-Freeman's " Field Arche-ology " as running from the Harroway near Long Sutton, north-east to Chobham,may be a continuation of this ridgeway.

33i

a Saxon road: for there is no getting away from the fact thatits characteristics are those of a summerway, not a ridgeway.Its route lies from Winchester along the north side of the Itchenvalley, along the modern road through the Worthys as far asItchen Stoke (though I am not sure of its exact course throughItchen Abbas); it then crossed the Itchen and went throughOvington and over the saddle of Tichborne Down to Bishop'sSutton; further, it coincides with the modern road as far asRopley Dean. It must then have gone straight on betweenNorth Street and Gilbert Street, and is visible again in Black-berry Lane, leading up to the watershed at Four Marks, whereit must have joined the ridgeway for a short distance, and thenit leaves it again on the right and proceeds to Alton along thenorth side of the Butts.

Now that so much of the parent ridgeway is established,this conception of the Pilgrims' Way as a summerway has noserious faults; it keeps as near to the ridgeway as the groundpermits, and adheres carefully to smooth slopes and easy gradients.

(xii.) I suggest as a possible modification that the portionthrough the Worthys and Itchen Stoke might be Saxon, andthat the true summerway left its parent on Gander Down, andwent down through Fulley Wood to the Itchen at the ford justnorth of Tichborne, only coinciding with Mr. Belloc's routeeast of that point. Certainly the track through Fulley Woodhas all the characteristics of age, though there is now a gapin its course on Gander Down. „

(xiii.) A branch left the South Hants ridgeway southwardfrom Millbarrow Down along the edge of Dur Wood: it waslater used as a Salt-way, and must have reached the coast byway of Upham and Durley, but beyond Dur Wood it is how lost..

(xiv.) Near the tumuli on Wether Down near East Meon,a branch left the South Hants ridgeway on the north ; it canbe followed over Salt Hill and Henwood Down, and crossed theMeon at Westbury. Here there is a fork, and it continues intwo parallels, one a lane north across the railway to Stock Farmand. Bailey. Green, the other further east past Peake Farm toPrivett. Beyond these points the ridgeway vanishes, but itmust have gone on past the Jumps and over West Tisted Commonto join road (X) near Hill Farm. Whether the eastern parallel,which, according to. Dr. Grundy, is shown by its straightness,and the name Straet given it in a Saxon charter, to have beenstraightened and metalled by the Romans between the Meonand Peake Farm, was the beginning of a summerway whichwent on down through East Tisted under the edge of the hillsto Alton, or was merely a bi-form which rejoined the westernparallel again near the Jumps, I am not sure: the " romaniza-tion " might be adduced in favour of either view.

(xv.) There is a group of roads leading north on the westbank of the Itchen, about which there is some doubt. From

332

Silkstead an old lane runs northward to Yew Hill: there aretraces of it as far south as Hocombe Bridge, and Isaac Taylorshows it coming from the neighbourhood of Toot Hill, near

• Southampton, by way of Chilworth. I t is not a ridgeway, butthere are many hollow tracks on Yew Hill, and the road whichgoes north thence past Oliver's Battery (now partially obliteratedby Pitt Corner Camp) looks like one, and must have joined theSouth Hants ridgeway on Teg Down or Sleeper's Hill, just aboveWinchester. Though there is a green road north-east fromHursley along the valley past the Kennels, there,is no vestigeof a way on the ridge between, south-west of the hollow trackson Yew Hill, and I am not sure where the ridgeway went.

(xvi.) From Yew Hill hollow tracks .lead down to theItchen, at Shawford, and beyond, White Lane,1 leading north-north-east up on to Twyford Down, marks the continuation ofthis branch. The track, after turning east and running alonga small fir plantation, ceases, but a deep hollow way can bemade out running up through the fields to the Winchester-Bishop'sWaltham road, just south of the second milestone, beyond whichthe southern portion of " Steeplechase" Lane shows how itwent on to join the South Hants ridgeway by the farm justsouth of Deacon Hill.

(xvii.) Whether any part of " King's Lane," which leadsover from Compton Down to St. Cross, is pre-Saxon, is doubtful,but there are well-marked trails connecting it with the hollowtracks of the ridgeway on Yew Hill, as well as a footpath andhollow trails which lead to it from the green road at HursleyKennels. On the north-west slope of St. Catherine's Hill thereis a bank, .and what appear to be very worn and shallow hollowtracks leading east-north-east across the hillside, and these maybe the relics of a continuation of " King's Lane " across theItchen to meet the South Hants ridgeway behind St. Catherine'sHill.

(xviii.) From Merdon Castle a track leads north and westalong Southlynch plantation: hollow tracks to the west of theearthwork seem to show it came from further south. I t wentpast Violet Hill on to Pitt Down, where it must have joined theSouth Hants ridgeway, though the last part of its course hasbeen obliterated by cultivation, and can only be found in IsaacTaylor's map. I am not positive whether it went on north tojoin road (iii), as shown on the map in Dr. Williams-Freeman's" Field Archaeology," but I have not found it in West Woods.The route further north is marked by a lane east of UpperSomborne, but it is doubtful, and if this way did continue to thenorthward, its course is much more likely to have been straightthrough West Woods along the parish boundary to join road (iv)at its bend between Sparsholt and Upper Somborne Wood.

1 The name points to use as a Salt-way in the Middle Ages.

333

(xix.) There must once have been a ridgeway leading eastfrom Harewood Forest, via Tidbury Ring and Popham Beacons,to the Harroway near Farleigh Wallop, but with the exceptionof a very dubious hollow way near Popham, I have failed to find any certain traces of it.

The present condition of an old road shows what treatmentit has received since the Saxon invasion : much has been ploughedout, and over cultivated ground a long hedge or footpath issometimes left to mark the line. Where the old road remainsas a lane through tilled ground, it was utilized by the Saxons,and narrowed down to its present width by the cultivation oneach side. Long strips of narrow field along a road show itwas formerly wider, and often indicate a ridgeway: the samemay be said of wide strips between a road and its hedges.Hollow tracks on the bare down cannot be taken as evidence ofa pre-Saxon road: in their present state they are frequentlymediseval (as on Twyford Down, east of St. Catherine's Hill), oreven later (as at Black Dick's Corner, on the direct Winchester-Alresford road). Sometimes parish boundaries follow Britishroads, but not quite so frequently as in the case of Roman high-ways. Where there is a ridgeway ascending out of a valley,or crossing from one valley to another, it goes up by a spur. I tis remarkable that a Saxon road will go up along the adjacentgully, but will coincide with the ridgeway on the crest, sinkinginto a gully again on the other side; this often results in thedisappearance of the ridgeway except on the top of the hill.This is so in Parhholt Wood, where a Saxon road comes up a gully from Braishfield, and on the top meets the South Hantsridgeway which has come up the spur from Michelmersh: alsowith the south branch of the South Hants ridgeway on Long-wood Warren, with the ridgeway on the north of BramdeanCommon and elsewhere.

An important feature of the British road system was theroads leading down to the coast (where Bitterne was the chiefport), but these are not now traceable very far south, and alsoI am sure there were originally more of them than are now tobe found. Most of these would be replaced by the southwardRoman roads from Winchester, just as the South.Hants ridgewaywas replaced by the Roman road to Old Sarum.for four mileswest of Winchester.